02_comments_number_analysis

grouped analysis

pronoun_by_post_num %>%
  group_by(post_num, total_num_com_in_sub) %>% 
  summarise(
    #agg_author = sum(num_authors),
    agg_i      = sum(num_authors*contains_i     )/sum(num_authors),
    agg_me     = sum(num_authors*contains_me    )/sum(num_authors),
    agg_we     = sum(num_authors*contains_we    )/sum(num_authors),
    agg_us     = sum(num_authors*contains_us    )/sum(num_authors),
    agg_they   = sum(num_authors*contains_they  )/sum(num_authors),
    agg_them   = sum(num_authors*contains_them  )/sum(num_authors),
    agg_you    = sum(num_authors*contains_you   )/sum(num_authors)
  ) %>%
ggplot() +
  geom_point(aes(post_num,agg_i,color=total_num_com_in_sub))

Error in summarise_impl(.data, dots) : 
  Evaluation error: object 'num_authors' not found.

Super Grouped results

all_pronouns <- pronoun_by_post_num %>%
#filter(post_num > 0, total_num_com_in_sub > 100), 
mutate(
  num_til_quit = total_num_com_in_sub - post_num
)
library(plotly)
ggplotly(
ggplot(filter(all_pronouns, num_til_quit < 10), aes(post_num,contains_me,group=num_til_quit, color = num_til_quit)) +
  #geom_point()+
  geom_smooth(se = FALSE)
)

`geom_smooth()` using method = 'loess' and formula 'y ~ x'

  ggplotly(
ggplot(filter(all_pronouns,total_num_com_in_sub < 100, num_til_quit > 6, post_num < 100)) +
  #geom_point(aes(post_num,contains_we,color=total_num_com_in_sub))+
  geom_smooth(aes(post_num,contains_i,color=total_num_com_in_sub))
)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

  
  ggplotly(
ggplot(filter(all_pronouns,total_num_com_in_sub < 50, post_num < 20)) +
  #geom_point(aes(post_num,contains_we,color=total_num_com_in_sub))+
  geom_point(aes(post_num,contains_we,color=total_num_com_in_sub))
)

  
  ggplotly(
ggplot(filter(all_pronouns,total_num_com_in_sub < 500, post_num < 200)) +
  #geom_point(aes(post_num,contains_we,color=total_num_com_in_sub))+
  geom_point(aes(total_num_com_in_sub,contains_we,z = post_num))
  #geom_smooth(aes(total_num_com_in_sub,contains_we,z = post_num),se = FALSE)
)

Ignoring unknown aesthetics: z

  filter(all_pronouns,total_num_com_in_sub < 500, post_num < 1000) %>%
  plot_ly(x=.$post_num, z=.$contains_i, y=.$total_num_com_in_sub, type="scatter3d", mode="markers", color = .$contains_we)

  
  filter(all_pronouns,total_num_com_in_sub < 500, post_num < 1000, num_til_quit < 300) %>%
  plot_ly(x=.$post_num, z=.$contains_we, y=.$num_til_quit, type="scatter3d", mode="markers", color = .$contains_we)

ggplotly(
ggplot(filter(all_pronouns,total_num_com_in_sub > 15, total_num_com_in_sub < 400,num_til_quit > 5)) +
  geom_smooth(aes(post_num,ZNcontains_we, color = "We")) +
  geom_smooth(aes(post_num,ZNcontains_i, color = "I")) +
geom_smooth(aes(post_num,ZNcontains_us,color = "Us"  ))+
geom_smooth(aes(post_num,ZNcontains_they,color = "They"))+
geom_smooth(aes(post_num,ZNcontains_them,color = "Them"))+
geom_smooth(aes(post_num,ZNcontains_you,color = "You" ))+
geom_smooth(aes(post_num,ZNcontains_me,color = "Me"  ))
  
)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'